ing unique and major resource institutions like these and be authorized to provide incremental funding to enable them to serve more people than their primary clientele. To achieve this, the Federal Government would offer to compensate such institutions for performing added services. The institutions, in turn, would have the option of accepting or rejecting a national responsibility for developing and sustaining their particular collections.

Charges may have to be levied for the use of some unique collections. When such compensation is required, appropriate fees and payment methods will need to be devised.

(3) To develop centralized services for networking. A similar responsibility of the Federal Government would be to sponsor and support centralized bibliographic and other services in the public and private sectors, when it can be reasonably demonstrated that such central services would benefit a majority of libraries and information centers or achieve economies of scale.

Examples of potential services include a national audiovisual repository, a national system of interlibrary communication, a national depository for the preservation of master microforms and "best copies" of all works of significant research value, or a national periodical bank. The National Library of Medicine (NLM), for example, is already the "library of final resort" for medical periodicals not held elsewhere in the country.

In the humanities there is a steady increase in the establishment of bibliographic data banks of abstracts of articles published in the United States and abroad. A need to combine these and other data banks for cross-disciplinary research in the humanities could lead to the development of a centralized computer retrieval service for the humanities available through the nationwide network.

The Commission, of course, recognizes that some functions are better performed locally than centrally. However, in many areas of the country some institutions are wastefully duplicating effort, performing repetitive processing, storing similar materials beyond those required to satisfy local everyday demands, and giving incomplete or limited services to the public because of the lack of centralized services. Existing national services that currently serve the

Program. Others that are needed would be initiated by the Federal Government.

The criteria to be followed in designating national collections and services, or in recommending their establishment de novo, will need to be carefully articulated in proposed legislation.

) To explore computer use. Computer technology is another very important part of the design of a future nationwide information network. Computers have become indispensable tools in the operation of library and information networks.

Today, libraries use computers for many phases of their operations: recording, control, dissemination, and retrieval of bibliographic information, catalog card production, circulation control, book ordering, serial records, and other routine library functions. Aside from these applications, research is also being pursued by libraries to find ways of using the computer to answer library reference questions; a number of libraries, in fact, have already begun to search computer bibliographic data bases. Libraries of the future might be expected to have the full text of certain materials stored in a form readable by machine.

The nationwide network may require several computer installations for centralized processing to help transform the machine-readable bibliographic records, produced by the Library of Congress and other national libraries, into forms (such as cards, book catalogs, special bibliographies, selective dissemination of information (SDI) services, etc.) suitable for decentralized use in each state. For each library, or each state, or that matter, to operate a large-scale computer installation would be prohibitively expensive in most cases. The cooperative, time-shared, multi-institutional approach to computer usage, supported by the Commission, appears to be the most economic and efficient solution.

Computer installations in the nationwide network would carry out three functions: the first, dedicated to bibliographic production (the processing of machine-readable

tapes produced by the national libraries into by-products required by the local institutions); the second, devoted to service uses (recording holdings, making referrals, managing interlibrary loans, searching data bases, performing interactive searches of bibliographic and abstract files, etc.); and the third, related to the management and accounting function of network operation, including inter-system payments of suppliers of information. The existence of several computer centers for interstate use in the network will not offset the need for some libraries to maintain their own computers-probably dedicated minicomputers-to satisfy local internal processing needs. In fact, the minicomputer may eventually become a distinct and direct functional component within a national communications and computer network. Computers at the multistate level of the national network would probably be a set of large, fast, time-shared computers, with transmitting and receiving terminals in the member institutions.

(5) To apply new forms of telecommunications. Since the main purpose of a nationwide network is to place the user in contact with his materials, finding ways of speeding up the delivery of information constitutes one of the more important aspects of the network concept. A nationwide network must incorporate appropriate means of communicating rapidly and effectively with the facility at which the desired material is located. It is in regard to the techniques which allow optimal interconnection between user and resource that the greatest change in current thinking and practices will be required.

Of all the different kinds of equipment used by libraries for interlibrary communications, the one which has received widest acceptance, other than voice-grade telephone, as a low-cost practical tool, is the teletypewriter. Teletype communications between and among libraries exist in both informal and formal network configurations. They are generally used to augment library holdings on a reciprocal basis, to provide for general communications with other libraries, to serve as a channel for querying union catalogs, and to accommodate reference questions and services. Business, industry and government also utilize teletype for exchanging information.

A future telecommunications system used for a nationwide information network will eventually need to integrate teletype, audio, digital, and video signals into a single system. This concept is an important aspect of the design of a modern communications system for information exchange. "Integrated telecommunications systems" have become practical only during the past few years, and commercial and governmental efforts are underway to provide these unified facilities on a large scale. Within the next few years, domestic communication satellites will be operating over the United States, thus further enlarging the nation's capability to exchange information in all forms.

Although distribution of documents from, say, holographic or microform collections through electrical channels to individual libraries, or even directly to the user, will soon be technically feasible, the bulk of information will, most probably for a long time to come, be transmitted over regular communication channels such as mail, parcel service, intercity bus, rail, dedicated interlibrary delivery systems, bookmobile, and other means. Even though, at the present time, many commercial telecommunication companies are upgrading their lines, it would appear that the regular costs for library and information telecommunications would still be too high, and that an exception to the Federal telecommunications regulations may be needed to guarantee reasonable rates for interstate information exchange.

The Commission believes that rapid and inexpensive telecommunication among members of the nationwide network could turn out to be the greatest boon ever to the national distribution of knowledge for education and progress. For this reason, the responsible agency would be directed to explore all possible avenues leading to reasonable communication rates for library and information networking purposes. First, as an interim step, the possibility should be explored of incorporating this type of communication into the normal Federal Telecommunications System (FTS). In this case, special legislation may be necessary to authorize interstate use of the FTS system free of charge or at a reduced rate. Second, approaches might be made to the Federal Communications Commission (FCC), or the National Aeronautics and Space Agency (NASA),

for permission to use satellite communication channels, at first for experimental purposes, and later for regular traffic. Or, alternatively, the cost of interstate communications could be borne by the institutions that use the network, either by being subsidized directly by the state or Federal Government, or through charges levied against the individual user.

Many European countries have already begun to provide communication links at lower tariff rates in order to influence and stimulate the development of national information systems.

The United States Government is in a position to give the whole nationwide library and information network an initial impetus by subsidizing low-cost rates until the entire scheme reaches a level of usage that ensures its economic viability.

(6) To support research and development. Transforming the nation's heterogeneous information facilities and services into a nationwide network will pose many new problems. Some of these problems will arise from the application of the new technology, some will derive from the effects of new information systems on users, and others will originate with the profession itself as it struggles with the dynamics of change.

A stronger Federal program of research and development, through grants and contracts, can provide an overall framework within which common investigations can be carried out. By concentrating specialized skills on crucial common problems, the Federal Government helps reduce duplicate and costly piecemeal research that would otherwise be performed by the states, provides for research and demonstration across jurisdictional boundaries and, at the same time, greatly accelerates the rate at which new methods and equipment can be transformed into operating systems.

For example, a Federal policy should be enunciated. which encourages and facilitates the development of telecommunication technology and services especially suited to interlibrary communication (and another Federal policy should encourage the sharing of relevant computer software developed at government expense).

The Commission believes that a vigorous Federal research and development effort is essential. At present, the